U.S. patent number 4,015,480 [Application Number 05/635,722] was granted by the patent office on 1977-04-05 for balancing method and apparatus.
This patent grant is currently assigned to Carl Schenck AG. Invention is credited to Alfred J. Giers.
United States Patent |
4,015,480 |
Giers |
April 5, 1977 |
Balancing method and apparatus
Abstract
The present balancing method measures instantaneous analog
unbalance representing signals and converts these signals into
digital signals. The conversion involves multiplying the measured
signals with reference values at least three times within each
period of a reference signal. Said reference values are sine or
cosine values of an oscillation having the same frequency as the
unbalance vibration. The sine and cosine reference values are taken
at points of time corresponding to the respective multiplication
point of time. The products of all individual multiplications are
separately summed for sine and cosine values and at least for one
period of the reference signal. Thereafter mean values are formed
from the two sums. The just mentioned signal processing steps are
performed by digital circuits including digital multipliers,
digital summing networks, digital mean value formation circuits, as
well as a clock signal generator.
Inventors: |
Giers; Alfred J. (Rossdorf,
DT) |
Assignee: |
Carl Schenck AG (Darmstadt,
DT)
|
Family
ID: |
5933467 |
Appl.
No.: |
05/635,722 |
Filed: |
November 26, 1975 |
Foreign Application Priority Data
|
|
|
|
|
Dec 14, 1974 [DT] |
|
|
2459282 |
|
Current U.S.
Class: |
73/462;
700/279 |
Current CPC
Class: |
G01M
1/225 (20130101) |
Current International
Class: |
G01M
1/00 (20060101); G01M 1/22 (20060101); G01M
001/22 () |
Field of
Search: |
;73/462 ;235/151.3 |
Foreign Patent Documents
Primary Examiner: Gill; James J.
Attorney, Agent or Firm: Fasse; W. G. Roberts; W. W.
Claims
What is claimed is:
1. A method for determining the unbalance of a rotating body to be
balanced and having an unbalance which causes an unbalance
oscillation having an unbalance frequency, comprising producing
instantaneous, unbalance representing signals, providing a
reference signal having a reference period, said reference signal
having a frequency equal to said unbalance frequency, producing
reference values at least at three points of time during said
reference period of the reference signal, said reference values
representing or approximating sine and cosine reference values with
respect to said reference period at said three points of time,
miltiplying in a digital manner said unbalance representing signals
with said sine and cosine reference values at said three points of
time to produce respective sine and cosine products, separately
summing in a digital manner said sine products at least during one
period of said reference signal to produce a sine sum, separately
summing in a digital manner said cosine products at least during
one period of said reference signal to produce a cosine sum, and
producing an unbalance indicating signal from said sine and cosine
sums by digitally forming means values from said sums.
2. The method according to claim 1, wherein the formed mean values,
which represent the unbalance components in the reference system
defined by the reference signal, are made available in digital form
for further use.
3. The method according to claim 1, wherein the unbalance
representing measured signals, which have superimposed thereon
noise components of varying noise frequencies, are multiplied
repeatedly with the respective reference values during the period
of the noise component having the highest noise frequency.
4. The method according to claim 1, wherein said digital
multiplying takes place at equal time spacings.
5. The method according to claim 1, wherein said sine reference
values and said cosine reference values approximate an exact sine
and cosine value to any desired degree of precision.
6. The method according to claim 1, further comprising varying the
time duration during which said summing and the mean value
formation take place in such a manner that a respective mean time
duration is equal to a period of the reference signal.
7. The method according to claim 1, wherein said summing is
performed within a summing time duration in such a sequence that
small intermediate sums are provided.
8. The method according to claim 5, wherein within any summing time
duration those multiplication values are added up which are spaced
from each other in time by one half revolution of the body to be
balanced.
9. The method according to claim 1, further comprising storing the
individual multiplication products produced within a summing time
duration in an intermediate memory.
10. The method according to claim 1, further comprising providing
the reference values in digital form.
11. The method according to claim 1, wherein the reference signal
is provided as a single pulse at the beginning of each period.
12. The method according to claim 1, further comprising providing
the reference signal in the form of a sine alternating voltage and
in the form of a cosine alternating voltage, and deriving said
reference values in digital form from said alternating voltages for
multiplication with the unbalance representing signals.
13. The method according to claim 1, comprising determining said
multiplication points of time by means of at least three pulses
which pulses are uniformly distributed over a period of the
reference signal.
14. The method according to claim 13, wherein said three pulses are
derived from pulse signal means such as a generator, or a frequency
multiplier, or a clock signal generator.
15. The method according to claim 1, further comprising attenuating
the unbalance representing measured signals to the control of
modulating range of the following circuits, prior to transforming
the measured signal into digital form.
16. The method according to claim 1, wherein the unbalance
representing measured signals have superimposed thereon extraneous
noise components of varying noise frequencies, said method
comprising attenuating noise components above a given frequency by
means of filter.
17. The method according to claim 1, further comprising storing
correction values in a memory and influencing the unbalance
representing signals by at least one correction value prior to
displaying or further processing.
18. The method according to claim 1, further comprising digitally
transforming the balancing signals which represents the unbalance
information into any desired reference system and displaying or
using the balancing signal in digital form.
19. An apparatus for producing an unbalance representing signal for
balancing a rotatable body to be balanced, comprising unbalance
signal pick-up means to provide periodic unbalance representing
signals, digital multiplier means having inputs and outputs, analog
to digital signal converting means operatively interconnecting said
pick-up means to said inputs of said multiplier means, digital
signal summing means connected to said outputs of said multiplier
means, digital mean value formation means connected to the outputs
of said signal summing means for providing said unbalance
representing signals, means providing reference values having
values dependent upon the instantaneous angular displacement of
said rotatable body, means applying said reference values to said
multiplier means for multiplying said unbalance representing
signals with said reference values, and timing means connected to
said multiplier means for effecting multiplication therein at least
at three predetermined points of time during a period of said
unbalance representing signals.
20. The apparatus according to claim 19, wherein said digital
multiplier means comprise digital memories for the unbalance
representing signals.
21. The apparatus according to claim 19, wherein said digital
multiplier means comprise a digital memory for said reference
values.
22. The apparatus according to claim 19, further comprising sample
and hold circuit means arranged upstream of the digital multiplier
means for the simultaneous ascertaining and storing of the measured
unbalance signals provided in analog form by said pickup means, as
well as for the reference values provided in analog form by said
timing means.
23. The apparatus according to claim 19, further comprising at
least one analog to digital converter arranged ahead of the
multiplier means for converting the signals measured by said
pick-up means in analog form and for converting the reference
values provided by said timing means in analog form.
24. The apparatus according to claim 23, further comprising
measuring point selector switch means arranged downstream of the
sample and hold circuit means and upstream of the analog to digital
converter.
Description
BACKGROUND OF THE INVENTION
The invention relates to a method and apparatus for balancing the
unbalance of rotational bodies, wherein the measured values are
ascertained and processed by means of electro-mechanical measuring
transducers, reference signal generators, and multiplication
devices.
In prior art balancing methods it is usual to ascertain the
unbalance electrically in two measuring planes. The forces caused
by the unbalance are transmitted to the bearings of the rotating
body to be balanced and respective unbalance signals are evaluated
and processed by electrical means. Load cells, displacement pick-up
means, and speed transducers may be used for ascertaining the
unbalance values which will be referred to hereafter as measured
signals or values. The measured signals provided by the detectors
or transducers normally comprise in addition to the unbalance
information a mixture of oscillations. These measured signals are
processed in analog electrical circuits. Such analog measured
signal processing normally provides values for the unbalance of the
body being tested by indicating the size and angular position of
the unbalance, or by indicating components thereof, mostly in two
measuring planes. The unbalance values, or rather the signals
representing the unbalance values may be transmitted relative to
correction planes through the rotor, they may be indicated,
registered and held ready for further processing or for controlling
the unbalance correction. The correction is normally accomplished
by the balancing machine, which automatically starts and controls
the correction procedure.
Conventional balancing methods operate relatively slowly especially
if complicated balancing problems are to be solved and if a high
precision in the balancing must be maintained. Thus, prior art
balancing methods may require a substantial length of time for
permitting the balancing. Substantial and complicated balancing
problems also require complicated electric circuits. The production
of such circuits and their operation may involve errors which
falsify the measured results or which may interfere with the
testing sequence. Another drawback of analog balancing devices is
seen in that it is not possible to directly connect these devices
to a central data gathering and data processing system, for
example, to control the mass production of rotors or the like.
German Patent Publication 1,298,735 discloses an apparatus wherein
the working tools for the unbalance correction are controlled by a
digital device. However, the unblance itself is ascertained and
processed by analog means.
OBJECTS OF THE INVENTION
In view of the foregoing, it is the aim of the invention to achieve
the following objects singly or in combination:
to provide a balancing method which delivers the unblance
information directly in digital form and without any detour through
an analog method;
to provide a balancing method which operates reliably at high speed
and with precision so that the method may be employed even for
complicated balancing requirements;
to simplify the operation of a balancing method to thereby
substantially avoid or reduce any operating or human errors;
to provide means for selecting reference values in such a manner
that the precision of the balancing method may be adapted to
different requirements;
to standardize the circuit blocks employed for the processing of
the measured values in connection with balancing machines so that a
simplification, as well as a substantial adaptability of the
balancing equipment is accomplished with regard to different
balancing problems; and
to provide a method and apparatus for balancing the unbalance or
rotary bodies, which method and apparatus are directly compatible
with a central electronic data processing system, for example, for
controlling and monitoring the production especially where the mass
production of rotating bodies is involved.
SUMMARY OF THE INVENTION
According to the invention the above objects have been achieved by
multiplying the measured instantaneous unbalance signals, which are
superimposed by undesired extraneous vibrations, at least three
times within the period of a reference signal with reference
values, whereby these reference values are the sine and cosine
values corresponding to the respective multiplication point of
time, and derived from an oscillation having the same frequency as
the unbalance oscillation. The products of the individual
multiplications are summed up separately as sine and cosine values,
at least over one period of the reference signal, whereupon a means
value is formed from the two separate sums, and all the individual
operations are performed in a digital manner.
The repeated multiplication of instantaneous values of the measured
signal withe reference signals, their separate summing as sine and
cosine values and the mean value formtion, whereby all these steps
are performed in digital manner has the advantage that the
unbalance information is directly available in digital form,
whereby high speed data processing for ascertaining the unbalance
may be employed. This feature is especially advantageous in
connection with automatically operating balancing machines and
could not be attained by prior art methods. Another advantage of
digitally ascertaining the unbalance information is seen in that
the precision of the balancing may now be easily adapted to any
particular requirements simply by selecting suitable method steps,
for example, by increasing the number of multiplications within a
summing time duration. By selecting the reference values
corresponding it is possible to adapt the balancing precision to
different requirements. Yet another advantage is seen in that even
complicated balancing problems which heretofore could be solved
only at substantial expense, may now be solved according to the
invention in a rapid and reliable manner. Furthermore, the
apparatus according to the invention may be realized in many
instances at lower costs and performing the present method is also
likely to be less expensive than prior art methods in view of the
technical advance achieved in the production of electronic circuit
elements and computer devices.
The apparatus for performing the present method comprises,
according to the invention, a digital multiplier, a digital summing
and mean value forming network, as well as a clock signal generator
for controlling the multiplication sequence of measured signals and
reference values. Such an arrangement comprising predominantly
digital circuit elements and circuit blocks, makes it possible to
standardize the data processing of measured values in balancing
machines, whereby simultaneously a substantial adaptability
relative to different balancing problems is provided. Another
advantage of the invention is seen in that several balancing
machines may be connected to an apparatus according to the
invention which may in turn cooperate with a central electronic
data processing system. Thus, the invention is particularly useful
for controlling and monitoring large scale manufacturing
operations. Advantageously, the multiplication device comprises
digital memory means for the measured values and for the reference
values. This feature minimizes the requirements for making the
reference values available. In addition, it is possible to store
any desired correction values.
Preferably, the apparatus according to the invention comprises
holding circuits arranged ahead of the multiplication device for
the simultaneous receiving and storing of the analog measured
signals from the measuring transducers and of the analog reference
values. The holding circuits make it possible to sequentially
convert the measured analog signals representing the unbalance and
the analog reference values into respective digital values. Thus, a
single analog to digital converter may be employed for this
purpose. Such an analog to digital converter will be arranged
upstream, as viewed in the direction of signal flow of the
multiplication device and downsteam of the holding circuits.
Preferably, a measuring point selector switch will be arranged at
the input of the A-D converter.
BRIEF FIGURE DESCRIPTION:
In order that the invention may be clearly understood, it will now
be described, by way of example, with reference to the accompanying
drawings, wherein:
FIG. 1 is a circuit block diagram illustrating the method steps
according to the invention in a circuit arrangement according to
the invention;
FIG. 2 illustrates a portion of the arrangement according to FIG. 1
with measuring selector switch means;
FIG. 3 illustrates the generation of the reference signals and of
the reference values by means of a pulse generator and a memory;
and
FIG. 4 shows an impulse generator suitable for practicing the
invention.
DETAILED DESCRIPTION OF PREFERRED EXAMPLE EMBODIMENTS:
The described, preferred embodiment refers to the balancing of
rigid rotors in two planes. All elements relating to the first
plane are provided with the index a. All elements relating to the
second plane are provided with the index b. However, the present
teaching may equally well be employed in connection with the
balancing in a single plane or in connection with the balancing in
more than two planes of a rotating body.
Referring to FIG. 1, the unbalance detectors 1a, 1b are connected
to a conventional balancing machine, not shown. These detectors or
pick-ups 1a, 1b may be of any conventional design. For example,
load cells or so called vibration pick-ups operating on the moving
coil principle would be suitable for the present purpose. The
output voltage of a moving coil vibration pick-up is directly
proportional to the vibration velocity. The output signals 2a, 2b
from the pick-ups in the form of alternating voltages are supplied
to attenuators 3a, 3b. These output signals represent the
unbalance. Normally these unblance representing signals have
superimposed thereon a noise component of varying frequencies. This
noise component represents unwanted extraneous oscillations. Such
attenuators are well known in the art, for example, so called
rheostats may be used. The attenuators may operate in steps or
stepless, that is in a continuous manner. The attenuators may be
adjusted by hand or automatically. For the latter purpose, the
output of the attenuators 3a, 3b is connected to a respective
signal follower circuit 4a, 4b. Such signal folower circuits are
known in the art and comprise normally a peak detector connected
with its output to two trigger or comparator circuits arranged in
parallel, one comparator being biased for high level detection,
whereas the other is biased for a low level detection and the
control output 4a', 4b' of each signal follower is connected to its
respective attenuator for the automatic adjustment of the latter in
response to the signal level provided at the output of the
attenuators. If the voltage at the output of the attenuator exceeds
a certain value, the respective signal follower will automatically
adjust its attenuator until the signal is again within a
predetermined level. Such level may be determined, for example,
with regard to the control range of the following circuit elements,
and the adjustment will always keep the signal amplitude within the
control range of the following circuit elements. In the illustrated
embodiment the signal followers 4a, 4b automatically adjust the
attenuators 3a, 3b respectively until any over-control of the
following circuit elements is eliminated. In a modification of the
shown circuit embodiment the signal followers 4a, 4b may be
inserted just downstream of the correction computer 18. A device
suitable for performing the functions of the correction computer 18
is known under the model number TMS 0117 NC manufactured by Texas
Instruments, Inc., Dallas, Tex., as a 10 digit decimal arithmetic
processor. Such a processor is illustrated in the respective block
78 in FIG. 1.
The signal follower circuits 4a, 4b are connected with their
outputs to analog filters 5a, 5b. Such filters are conventional and
are, for example, described in "Electronic Designers Handbook",
Section 16, McGraw-Hill Co., 1957. If desired, the filters 5a, 5b
may be inserted between the attenuators and the signal followers.
Under certain circumstances the filters may not have to be used at
all. Their use will depend on the size of the noise ratio. If the
following circuit means are not designed to process any noise
ratio, it is advisable to use the respective filters. Thus, for
example, where the noise components comprise rather high
frequencies a low pass filter would be used. However, no specific
requirements must be met by the frequency response characteristic
of the filter, because it is quite easy to correct any filter
transmission factors in connection with a correction computation to
which the signal will be subjected in its processing. For example,
force measuring balancing machines, which are so called "hard
bearing" balancing machines, involve high frequency noise
components.
The preprocessed measured voltages at the outputs 6a, 6b are
supplied to the inputs of an analog to digital converter (A-D
converter) 7a, 7b. The converter converts the analog measured
signal including the useful unbalance component and the noise
component into digital values at points of time which may be
determined by a clock signal generator forming part of the A-D
converter. It is also possible to connect the clock signal
generator to a reference signal generator, for example, reference
signal generator 9, to produce a predetermined number of pulses for
each revolution of the body to be balanced. The operation of the
reference signal generator 9 is controlled by the rotation of the
body to be balanced. For simplicity's sake the connection between
the rotating body to be balanced and the reference and clock signal
generators is not shown. Each pulse produced by the clock signal
generator causes an analog to digital conversion of the
instantaneous analog values into respective digital values, as will
be described in more detail below with reference to FIG. 3.
Suitable analog to digital converters for the present purpose are
described in "Analog-Digital Conversion Handbook", published by
Analog Devices Inc., Norwood, Mass. 02062. The reference signal
generator 9 may comprise, for example, a two phase sine wave
generator of a type customary in balancing equipment and driven by
the rotor to be balanced as mentioned above. Instead of producing a
sine voltage the reference signal generator may produce one or
several pulses for each revolution of the body to be balanced. In
any event, the reference signal generator produces a signal by
means of which the angular position of the body to be balanced is
fixed.
In order to determine the unbalance of a rotational body, the
measured voltages are multiplied by sine and cosine analog
reference voltages when the so called multiplying method is
employed. According to the invention, the instantaneous values of
the measured voltages are multiplied with reference values. These
reference values are the digital sine and cosine values which
correspond to the respective multiplication point of time. In the
alternative, the reference values may approximate the just
mentioned sine and cosine values to any desired degree and such
approximation reference values may be derived from an oscillation
corresponding in its frequency to that of the unbalance
oscillation.
The reference signal generator 9 produces in the illustrated
embodiment a sine reference voltage sin-u and a cosine reference
voltage cos-u. These reference voltages are employed as reference
signals and are simultaneously used for producing the reference
values. The two voltages sin-u and cos-u are supplied to analog to
digital converters (A-D) 10.1 and 10.2 where they are converted
into respective digital values simultaneously with the conversion
of the measured voltages in the converter 7a, 7b. The point of time
at which the conversion takes place are determined by the clock
signal generator (not shown), whereby at least three conversions
are to take place for each period of the reference signal. In other
words, for each revolution of the body to be balanced, at least
three conversions will take place.
The output teminals of the digital converters 7a, 7b, 10.1 and 10.2
are connected to respective inputs of a multiplier 11, which is
part of a data computer 20. Such a multiplier, may for example, be
one known under the model no. TMS 0117 NC, which multiplies the
measured values with the reference values. The combination of
measured values and reference values is such that the instantaneous
digital measured value for each measuring plane is multiplied with
the corresponding digital sine and cosine reference value, so that
four products are provided for the two measuring planes. The block
diagram of FIG. 1 signifies the four multiplication values by
providing respective four channels, namely, 11a1, 11a2, 11b1, and
11b2. Channels 11a1 and 11a2 correspond to the first measuring
plane a and channels 11b1 and 11b2 correspond to the second
measuring plane b. Which may be of the type No. P 2100 by
Hewlett-Packard
Each measured value is multiplied with the sine reference values in
channels 11a1 and 11b1 whereas each measured value is multiplied
with the cosine reference values in channels 11a2 and 11b2. The
outputs of the multipliers are connected to respective intermediate
storage or memory means 12 where the values are summed over a
summing period. The processor 20 of the above mentioned type
includes such intermediate summing memory means. The four sums are
supplied from the memory to respective mean value computing
circuits 13. The means value is computed by dividing the sums by
the number of multiplications performed during each summing time
period or by the number of sensings taken during each summing
period or summing duration.
According to the invention, at least three measuring signals are
sensed or three multiplications are performed for each period of
the reference signal in order to ascertain the size and angular
position of the unbalance vibration or oscillation. However, in
order to increase the precision of the operation, the number of
sensings or rather the number of multiplications per period may be
increased as desired. The increasing of the multiplications is also
a suitable expedient to compensate for a non-uniform distribution
of the sensings or rather multiplications over a period of the
reference signal. This non-uniform distribution may, for example,
occur when the reference signal from the reference signal generator
9 and the signal from the clock generator in the A-D converter 7a,
7b are not precisely synchronized with each other. For the same
reason it may be advantageous to produce the mean value over a
number of periods which is as large as possible.
A number of sensings and a correspondingly large number of
multiplications per period of the reference signal is also
desirable when it is intended to eliminate noise vibrations having
higher frequencies. In order to suppress a noise vibration of a
given frequency, it is necessary that at least two sensings take
place for each period of such noise vibration.
According to a further feature of the invention the mean values
need not be produced over a single period or an integral multiple
of a single period, but rather the duration of the time during
which the summing and mean value formation takes place may be
varied, whereby the mean value of the total time duration
corresponds to one period of the reference signal. This type of
mean value formation may be performed in a manner similar to that
described above. Further, it may be advantageous not to form the
arithmetic mean value but rather to perform the computation with
the smallest error squares.
As a result of the multiplication and of the mean value formation
there are now available four digital values which represent the
components of the vibration having a frequency corresponding to the
r.p.m. and relating to the respective measuring plane, whereby the
components are represented in a rectangular coordinate system.
These four digital values are supplied to a so called plane
computer 14 which is also part of the data processor 20. It is the
purpose of the plane computer 14 to compute from the two
oscillation components the two rotor unbalance components relative
to given correction planes in the rotor. Such computation may be
accomplished in known manner, for example, in connection with force
measuring balancing machines, whereby the spacings between the
measuring planes and the correction planes are taken into account.
The respective parameters may, for example, be supplied through a
keyboard to suitable peripheral devices or the parameter may be
made available in any other suitable manner for the plane
calculation, for example, from suitable memories. Subsequent to the
computation, the unbalance components are available in digital form
and related to two given correction planes. If the apparatus
according to the invention is employed in connection with a
displacement measuring balancing machine "soft bearing machine", it
is necessary to empirically adjust the plane computer.
Such empirical adjustment may easily be accomplished by employing
initially for the adjustment a balanced rotor or by simulating a
balanced rotor. For the adjustment a so called erasing or
compensating computer 15 is arranged between the mean value
computer 13 and the plane computer 14. The erasing computer has
four channels which deduct during the adjustment a digital value
from the value appearing at the output of the mean value computer
13 so that the output signal will be zero. The erasing pulses may,
for example, be provided by an erasing or compensator signal source
16 which derives the erasing signals or pulses from the digitized
reference signals.
For most practical purposes, it is not necessary to ascertain the
unbalance itself, rather it is sufficient to ascertain the size of
the required correction, in other words, to ascertain the
respective mass relative to a balancing or correction radius. Thus,
the output of the plane computer 14 is connected to a radius
computer 17, which in turn is, for example, connected to a memory
19 or which may be controlled by the operator manually to provide
for each correction plane a radius value. The radius computer 17
divides the measured values by the radius values and multiplies it
with a scale factor. As a result, one obtains the components of the
balancing masses for the two balancing or correction planes
directly in the units determined by the scale factor. The scale and
radius values may be supplied by the memory 19 or by adjustment
through the operator.
In many instances it is desirable to influence the unbalance signal
prior to its display or prior to its further processing. Thus, the
unbalance signal may be passed through a correction computer 18,
which is also a part of the signal or data computer 20 mentioned
above. The correction computer may take into account especially
errors inherent in the mechanical or measuring components of the
balancing machine, or which are inherent in the system. Most of
these errors will involve frequency responsive elements. Such
errors are measured once and stored, for example, in the memory 19.
Such errors may, for example, include the fact that the measured
voltage of the vibration pick-up means is frequency dependent
although the unbalance is the same. For example, the output voltage
of load cells or force measuring means is frequency dependent is a
square manner. In this instance, the r.p.m. of the body to be
balanced is measured by means not shown and a correction value
corresponding to the measured r.p.m. is called up or polled from
the memory 19 for corresponding correcting the result. Another
example to be mentioned here is phase or shifts errors of the
vibration pick-up means. In this instance, the measured value is
also corrected by a correction value which depends on the measured
r.p.m. and which may be stored for instance in the memory 19. The
frequency characteristic of the filters 5a, 5b may also be
corrected as mentioned above. This inverse frequency characteristic
may be stored in the memory and employed as an r.m.p. dependent
correction value. Further correction values may taken into account
the measuring at speeds near the mechanical resonance of the body
to be balanced. In this instance the increase of the mechanical
amplitude measured in hard bearing machines may be taken into
account.
Furthermore, certain corrections may be made which are not
frequency proportional. An example of this nature is the adjusted
position of the attenuator 3a, 3b. Thus, the digital values
supplied to the input of the correction computer 18 may be
multiplied by a value which is inverse to the adjustment position
of the attenuator. In this manner it is possible to eliminate
errors which might have been introduced through the
attenuators.
The result of the above described method steps comprises the two
components of the correction masses appearing in digital form at
the outputs 20, 21, 22, 23 and 24 of the correction computer 18.
These digital values are referenced relative to the two given
correction planes and relative to the given correction radii. In
addition, these values are provided in a rectangular coordinate
system. Those signals or digital values may now be further
processed in additional digital computing devices. For example, a
recomputation into a different coordinate system, such as a polar
coordinate system may be performed, and the values or signals may
be employed for a control function. After recomputing these values
into a polar coordinate system, they are available in their size
and angular components. These values may be digitally displayed or
they may be reconverted and displayed in an analog manner.
Recomputation may also be performed into any other coordinate
system, for example, for the purpose of a localized balancing
(unbalance correction at defined points). Another recomputation may
involve more than two correction planes having respective sectors
designed for the unbalance correction. Further, these values may
directly control the balancing machine to perform the correction
procedure. In another further computation it is possible to take
into account the eccentricity of an auxiliary shaft on which the
body to be balanced may be supported. Thus, the eccentricity of the
auxiliary shaft is deducted from the unbalance value. Yet another
computation may involve adapting the output values to a suitable
display scale, for example, the display may be on a logarithmic
scale, and the recomputation would take the largest output
component into account.
As a practical matter, the above described operations and devices
starting with the multiplier 11 and the correction computer 18,
with the exception of the erasing signal source 16, will be
combined in a digital data computer 20. However, the described
steps may be performed in a manner deviating from the described
example. Thus, techniques well known in digital data processing may
be employed, for example, a time multiplexing may be used including
sequence programs and so forth. In addition, it is possible to
perform the method steps related to the devices 14, 17, and 18 in a
different sequence. For example, the correction computing may be
performed subsequent to the formation of the means value and prior
to the plane and radius calculation.
Where the reference values, as described with reference to FIG. 1
are sine and cosine voltages sin-u, cos-u which must be converted
from their analog form into digital values, it is practical to
employ a measuring point selector switch. With the aid of such a
switch it is possible to use but one analog to digital converter
for converting the measured values, as well as the reference
values. A circuit diagram of such an arrangement is illustrated in
FIG. 2.
Referring to FIG. 2 the measured voltages 6a, 6b and the reference
voltages sin-u cos-u are first supplied to four hold circuits 30,
31, 32, and 33. These hold circuits or so called sample and hold
circuits may, for example, be realized by model SHC 23 or model SHC
85 produced by Burr-Brown, Tuscon, Ariz. These sample and hold
circuits 30, 31, 32, 33 ascertain simultaneously four instantaneous
values of the measured and reference voltages, and store these
values during the converting time. A measuring point selector
switch 34 supplies the respective instantaneous values to an A-D
converter 35 which converts the values sequentially from analog
into digital form. Such electronic selector switches 34 are, for
example, described in "Application of Operational Amplifiers",
McGraw-Hill Book Co., page 115, FIG. 4.5. The control of the
sequential sensing of the hold circuits through the selector switch
34 and the sequential conversion is controlled through a clock
generator 36 and a computer 37. This computer 37 also controls the
feeding of the digitial values to the multiplier 11. The further
steps are the same as described above with reference to FIG. 1. A
suitable clock pulse generator for realizing the generator 36 is
described, for example, in the just mentioned book "Application of
Operational Amplifiers", pages 154 and 155, FIG. 5.8. The computer
37 may, for instance, be of the type HP 2100 manufactured by
Hewlett-Packard.
FIG. 3 illusrates the generation of the reference signals and the
reference values through a pulse generator 40 and a memory 43. The
pulse generator 40 may be of the similar type as described above
relative to generator 9. The pulse generator 40 provides for each
revolution of the body to be balanced a single output pulse 41
which determines the begin and the end of a period corresponding to
one revolution. Thus a phase reference is established. A further
output 42 of the pulse generator 40 provides n additional pulses
uniformly distributed over the duration of one revolution or rather
over the duration of one period. A memory 43 has stored therein
sine and cosine values. This memory may be part of a computer 44
which in turn may, for example, be of the same type as mentioned
above with reference to the computer 37. The digital sine and
cosine values corresponding to each of the n pulses from the output
42 of the generator 40 are polled from the memory 43 in response to
these n pulses determining respective points of time. These sine
and cosine digital values are supplied to the respective multiplier
stages in the multiplier 11 forming part of the computer 44. At the
same time and in synchronism with the n pulses, the analog measured
values 6a, 6b are converted in the A-D converter 7a, 7b and
supplied to the respective multiplier stages, where these values
are multiplied with the corresponding digital sine and cosine
reference values. The pulse generator 40 constitutes with its
output 42 a clock signal generator which controls the sensing or
measuring of the analog values 6a, 6b as well as the converting of
these analog values into digital values, it also controls the
polling of the sine and cosine values from the memory 42 and thus
the multiplication of the measured values with the reference
values. The further processing of the individual values takes place
as described above with reference to FIG. 1.
Incidentally, the pulse generator 40 could be realized as shown in
FIG. 4 wherein an apertured disk 45 is driven by the same shaft as
the body to be balanced. A single light source 46 sends its light
through an aperture 47 to be sensed by a photocell 48, which in
turn actuates a trigger circuit 49, the output of which produces a
single pulse 41 for each revolution of the disk 45 and thus of the
body to be balanced. Similarly, a light source 50 sends its light
through a plurality of apertures 51 to be sensed by a photocell 52
to actuate a trigger circuit 53 which produces at its output the
pulse series 42 constituting the clock signal source as mentioned
above. The memory 43 could be read only memory of the type IM 5600
manufactured by Intersil, Inc., Cupertino, Calif.
Depending on the desired quality of the measured result, one will
determine the degree of approximation of the reference values
relative to the precise sine and cosine values. As mentioned, for
example, it is possible to employ instead of true sine and cosine
values of square wave function or a staircase function.
Instead of using a disk with a plurality of holes for producing the
clock pulses 42, it would be possible to employ a frequency
multiplier connected to the output 41 of the signal generator 40.
Such frequency multipliers are well known and may especially be
used in connection with balancing revolutions per minute in a
middle or high range. On the other hand, where the balancing speed
revolutions per minute) is rather low, it is suitable to employ
pulse generators, such as vibration pick-up directly secured to the
body to be balanced for controlling the sensing of the measured
values.
Another possibility of synchronizing the sensing with the polling
of the reference values from the memory 43 could employ a clock
signal generator such as 36, shown in FIG. 2, which in connection
with the computer 44, would synchronize the A-D converter 7a, 7b
and the polling of the reference values from the memory 43.
Incidentally, the erase signal source could, for instance, be
realized by means of a coding-switching arrangement manufactured as
model T 60-02 by Cherry Products, Corp. Highland Park, Ill.
Summarizing the foregoing disclosure of the invention, the mean
values representing the unbalance components in the reference
system determined by the reference signal, are made available
according to the invention in digital form for the further
processing. Such further processing may involve the digital display
of the values, their storage on their recording. By repeatedly
multiplying the measured unbalance signal, on which are
superimposed extraneous undesirable vibrations, with corresponding
reference values, within the period of the extraneous vibrations
having the highest frequency, it is possible to eliminate from the
measured signal the respective noise proportion, which periodically
occurs with said highest frequency, Preferably, the measured
unbalance signal is multiplied with the reference signal at points
of time equally time spaced from each other.
The sine and cosine values used for the multiplication may
approximate respective exact sine and cosine values to any desired
degree. Thus, reference values derived from a rectangular waveform
or from a stepped waveform may be used, whereby the multiplication
and means value formation is simplified and the respective costs
are reduced. The invention also suggests to vary the time duration
during which the summation and mean value formation takes place,
for ascertaining the measured value. Preferably, the mean value of
said time duration during which the measured values are obtained
should correspond to one period of the reference signal. This
feature has the advantage that non-periodic noise components of the
measured signal may be suppressed.
By making the individual intermediate sums of a summing sequence
small, the required storage capacity in the computer may be
reduced. Similarly, the required storage capacity may be reduced if
during a summing sequence those multiplication values are added to
each other which are spaced from each other by one half revolution
of the body to be balanced. In this connection it is practical to
store the individual multiplication products of one summing
sequence or duration, in an intermediate memory.
It is also advantageous to provide the reference values in digital
form. To this end analog values may be converted into digital
values or the reference values may be generated in digital form
initially. The digital reference values may also be made available
in a polling mode of operation. Preferably the reference signal is
supplied in the form of a single pulse per period, which pulse
determines the beginning of the period and the end of the period.
The reference signal may be supplied in the form of a sine
alternating voltage and of a cosine alternating voltage whih
constitute reference values and which are transformed into digital
values for multiplication with the measured unbalance signal.
The points of time at which multiplication shall take place are
predetermined by at least three reference signals which are
uniformly spaced from each other over the multiplication sequence.
Suitably, these pulses may be produced by an impulse generator
and/or a frequency multiplier, or these pulses may be derived from
a clock signal generator. By properly selecting the number of
multiplications which shall take place within a summing sequence or
within a summing period, it is possible to select any desired
degree of precision of the measured results. Preferably, the
measured signal is adapted or attenuated to the control or
modulation range of the circuits through which the measured signal
is to be processed, prior to transforming the measured signal into
digital form. Such adaptation avoids an overriding or over-control
of the signal processing circuits whereby possible measuring errors
are eliminated.
Noise or extraneous undesired vibrations of higher frequency in the
measured signal may be attenuated by filter means. This has the
advantage that the ratio between noise and useful signal components
may be kept small.
It is also practical to influence the unbalance information prior
to its display or prior to its further processing, by at least one
correction value which may be polled from a memory. Such correction
of the measured value is especially effective to take errors into
account which are inherent in the system, for example, in the
mechanical or in the measuring portion of the balancing machine.
Such errors which are inherent in the balancing system may, for
example, involve measured voltages, which have a frequency
depending characteristic, a phase error in the oscillation pickup
means, the frequency characteristic of filters, or an increase in
the mechanical amplitudes when approaching resonance and the
like.
By digitally transforming the unbalance information into any
desired reference system, it is possible to transform the
information into correction planes or correction radii, thereby
facilitating the control of the correction procedure. Although the
invention has been described with reference to specific example
embodiments, it is to be understood, that it is intended to cover
all modifications and equivalents within the scope of the appended
claims.
* * * * *